vecna
/
troll-patrol
mirror of https://git-crysp.uwaterloo.ca/vvecna/troll-patrol.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Model both accessible bridge and blocked bridge in analysis

This commit is contained in:
Vecna 2024-04-26 15:09:30 -04:00
parent b2eb244757
commit 31b27a291a
2 changed files with 295 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

110
src/analysis.rs
View File

@ -1,7 +1,7 @@
use crate::{BridgeInfo, BridgeInfoType};
use lox_library::proto::{level_up::LEVEL_INTERVAL, trust_promotion::UNTRUSTED_INTERVAL};
use nalgebra::DVector;
use statrs::distribution::{Continuous, MultivariateNormal};
use statrs::distribution::{Continuous, MultivariateNormal, Normal};
use std::{
cmp::min,
collections::{BTreeMap, HashSet},
@ -302,22 +302,52 @@ impl Analyzer for NormalAnalyzer {
let alpha = 1.0 - confidence;
let (mean_vec, sd_vec, cov_mat) = Self::stats(&[bridge_ips, negative_reports]);
let bridge_ips_mean = mean_vec[0];
let negative_reports_mean = mean_vec[1];
let bridge_ips_sd = sd_vec[0];
let negative_reports_sd = sd_vec[1];
// Artificially create data for alternative hypothesis
let num_days = bridge_ips.len() as usize;
let mut bridge_ips_blocked = vec![0; num_days];
let mut negative_reports_blocked = vec![0; num_days];
let bridge_ips_deviation = (2.0 * bridge_ips_sd).round() as u32;
for i in 0..num_days {
// Suppose bridge stats will go down by 2 SDs
bridge_ips_blocked[i] = if bridge_ips_deviation > bridge_ips[i] {
0
} else {
bridge_ips[i] - bridge_ips_deviation
};
// Suppose negative reports will go up by 2 SDs
negative_reports_blocked[i] =
negative_reports[i] + (2.0 * negative_reports_sd).round() as u32;
}
let (mean_vec_blocked, _sd_vec_blocked, cov_mat_blocked) =
Self::stats(&[&bridge_ips_blocked, &negative_reports_blocked]);
let mvn = MultivariateNormal::new(mean_vec, cov_mat).unwrap();
let pdf = mvn.pdf(&DVector::from_vec(vec![
bridge_ips_today as f64,
negative_reports_today as f64,
]));
if pdf < alpha {
(negative_reports_today as f64) > negative_reports_mean + negative_reports_sd
|| (bridge_ips_today as f64) < bridge_ips_mean - bridge_ips_sd
} else {
false
}
let mvn_blocked = MultivariateNormal::new(mean_vec_blocked, cov_mat_blocked).unwrap();
let pdf_blocked = mvn_blocked.pdf(&DVector::from_vec(vec![
bridge_ips_today as f64,
negative_reports_today as f64,
]));
// Also model negative reports in isolation
let nr_normal = Normal::new(negative_reports_mean, negative_reports_sd).unwrap();
let nr_pdf = nr_normal.pdf(negative_reports_today as f64);
let nr_normal_blocked = Normal::new(
negative_reports_mean + 2.0 * negative_reports_sd,
negative_reports_sd,
)
.unwrap();
let nr_pdf_blocked = nr_normal_blocked.pdf(negative_reports_today as f64);
(pdf / pdf_blocked).ln() < alpha || (nr_pdf / nr_pdf_blocked).ln() < alpha
}
/// Evaluate invite-only bridge with lv3+ users submitting positive reports
@ -339,25 +369,67 @@ impl Analyzer for NormalAnalyzer {
let (mean_vec, sd_vec, cov_mat) =
Self::stats(&[bridge_ips, negative_reports, positive_reports]);
let bridge_ips_mean = mean_vec[0];
let negative_reports_mean = mean_vec[1];
let positive_reports_mean = mean_vec[2];
let bridge_ips_sd = sd_vec[0];
let negative_reports_sd = sd_vec[1];
let positive_reports_sd = sd_vec[2];
// Artificially create data for alternative hypothesis
let num_days = bridge_ips.len() as usize;
let mut bridge_ips_blocked = vec![0; num_days];
let mut negative_reports_blocked = vec![0; num_days];
let mut positive_reports_blocked = vec![0; num_days];
let bridge_ips_deviation = (2.0 * bridge_ips_sd).round() as u32;
let positive_reports_deviation = (2.0 * positive_reports_sd).round() as u32;
for i in 0..num_days {
// Suppose positive reports will go down by 2 SDs
positive_reports_blocked[i] = if positive_reports_deviation > positive_reports[i] {
0
} else {
positive_reports[i] - positive_reports_deviation
};
// Suppose bridge stats will go down by 2 SDs
bridge_ips_blocked[i] = if bridge_ips_deviation > bridge_ips[i] {
0
} else {
bridge_ips[i] - bridge_ips_deviation
};
// Suppose each user who would have submitted a positive report but
// didn't submits a negative report instead.
negative_reports_blocked[i] =
negative_reports[i] + positive_reports[i] - positive_reports_blocked[i];
}
let (mean_vec_blocked, _sd_vec_blocked, cov_mat_blocked) = Self::stats(&[
&bridge_ips_blocked,
&negative_reports_blocked,
&positive_reports_blocked,
]);
let mvn = MultivariateNormal::new(mean_vec, cov_mat).unwrap();
if mvn.pdf(&DVector::from_vec(vec![
let pdf = mvn.pdf(&DVector::from_vec(vec![
bridge_ips_today as f64,
negative_reports_today as f64,
positive_reports_today as f64,
])) < alpha
{
(negative_reports_today as f64) > negative_reports_mean + negative_reports_sd
|| (bridge_ips_today as f64) < bridge_ips_mean - bridge_ips_sd
|| (positive_reports_today as f64) < positive_reports_mean - positive_reports_sd
} else {
false
}
]));
let mvn_blocked = MultivariateNormal::new(mean_vec_blocked, cov_mat_blocked).unwrap();
let pdf_blocked = mvn_blocked.pdf(&DVector::from_vec(vec![
bridge_ips_today as f64,
negative_reports_today as f64,
positive_reports_today as f64,
]));
// Also model negative reports in isolation
let nr_normal = Normal::new(negative_reports_mean, negative_reports_sd).unwrap();
let nr_pdf = nr_normal.pdf(negative_reports_today as f64);
// Note we do NOT make this a function of positive signals
let nr_normal_blocked = Normal::new(
negative_reports_mean + 2.0 * negative_reports_sd,
negative_reports_sd,
)
.unwrap();
let nr_pdf_blocked = nr_normal_blocked.pdf(negative_reports_today as f64);
(pdf / pdf_blocked).ln() < alpha || (nr_pdf / nr_pdf_blocked).ln() < alpha
}
}

205
src/tests.rs
View File

@ -1119,7 +1119,210 @@ fn test_analysis() {
);
}
// TODO: More tests
{
let mut date = get_date();
// New bridge info
let mut bridge_info = BridgeInfo::new([0; 20], &String::default());
bridge_info
.info_by_country
.insert("ru".to_string(), BridgeCountryInfo::new());
let analyzer = analysis::NormalAnalyzer::new(5, 0.25);
let confidence = 0.95;
let mut blocking_countries = HashSet::<String>::new();
// No data today
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
for i in 1..30 {
// 9-32 connections, 0-3 negative reports each day
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
8 * (i % 3 + 2),
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
i % 4,
);
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
}
// Data similar to previous days:
// 24 connections, 2 negative reports
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
24,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
2,
);
// Should not be blocked because we have similar data.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
// Data different from previous days:
// 104 connections, 1 negative report
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
104,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
1,
);
// This should not be blocked even though it's very different because
// it's different in the good direction.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
// Data different from previous days:
// 800 connections, 12 negative reports
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
800,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
12,
);
blocking_countries.insert("ru".to_string());
// The censor artificially inflated bridge stats to prevent detection.
// Ensure we still detect the censorship from negative reports.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
}
{
let mut date = get_date();
// New bridge info
let mut bridge_info = BridgeInfo::new([0; 20], &String::default());
bridge_info
.info_by_country
.insert("ru".to_string(), BridgeCountryInfo::new());
let analyzer = analysis::NormalAnalyzer::new(5, 0.25);
let confidence = 0.95;
let mut blocking_countries = HashSet::<String>::new();
// No data today
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
for i in 1..30 {
// 9-32 connections, 0-3 negative reports each day
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
8 * (i % 3 + 2),
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
i % 4,
);
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
}
// Data similar to previous days:
// 24 connections, 2 negative reports
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
24,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
2,
);
// Should not be blocked because we have similar data.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
// Data different from previous days:
// 104 connections, 1 negative report
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
104,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
1,
);
// This should not be blocked even though it's very different because
// it's different in the good direction.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
// Data different from previous days:
// 0 connections, 0 negative reports
date += 1;
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::BridgeIps,
date,
0,
);
bridge_info.info_by_country.get_mut("ru").unwrap().add_info(
BridgeInfoType::NegativeReports,
date,
0,
);
blocking_countries.insert("ru".to_string());
// This should be blocked because it's different in the bad direction.
assert_eq!(
blocked_in(&analyzer, &bridge_info, confidence, date),
blocking_countries
);
}
// TODO: Test stage 3 analysis
}